abdin2024-phi3

arXiv: 2404.14219

TLDR (English)

3.8B Phi-3-mini approaches GPT-3.5 on multiple benchmarks; continues to validate "high-quality synthesis + curriculum data" Phi recipe. Representative work for edge/local large models.

TLDR（中文）

3.8B Phi-3-mini 在多个评测上接近 GPT-3.5；继续印证"高质量合成 + 课程数据"的 Phi 配方。是端侧/本地大模型的代表工作。